본문 바로가기
  • Home

Design of an Automated Framework for Applying Generative AI-Based Source Code Obfuscation Techniques

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2025, 30(1), pp.73-85
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : December 9, 2024
  • Accepted : January 2, 2025
  • Published : January 31, 2025

Jihun Han 1 Seung-A Park 1 Joonseo Ha 1 Chang-min Lee 2 Kyung-mi Jung 1 Mee Lan Han 1 Jun-Seob Kim 1 Geumhwan Cho 1

1고려대학교
2고려대학교세종캠퍼스

Accredited

ABSTRACT

Source code obfuscation is an essential technique for software security and intellectual property protection. Traditional source code obfuscation methods depend on human-driven processes or predefined algorithms implemented by obfuscation tools. As a result, it becomes difficult to effectively manage the quality and complexity of obfuscated code. However, given that the required level of obfuscation differs based on the threat model, we need to develop techniques that can flexibly adjust the level and complexity of obfuscation. This paper proposes an automated framework that generates obfuscated source code from original source code using GPT-4o. The proposed framework generates prompts based on 12 obfuscation techniques and utilizes these prompts as input to GPT-4o. The generated obfuscated source code is executed and verified to ensure that it maintains the same intended functionality as the original code. Experimental results demonstrate that the proposed framework successfully executed 48 out of 60 obfuscated source codes while effectively applying the intended obfuscation techniques.

Citation status

* References for papers published after 2023 are currently being built.

This paper was written with support from the National Research Foundation of Korea.